The S&P 500 was higher 12 months after the first rate cut in 7 of the 9 Federal Reserve easing cycles that began between 1984 and 2024, with an average price gain of 8.7%. That average blends two different histories. Where no recession began within 2 years of the first cut, the index rose every time, by 16.7% on average. Where one did, it averaged a 1.2% loss, and the only 12-month declines followed the cuts of January 2001 and September 2007.
The study finds each cycle in the Fed's own target series on FRED, measures the index with Robert Shiller's monthly S&P 500 data through October 2026, and dates recessions from the chronology kept by the National Bureau of Economic Research (NBER). Each of those choices moves the numbers, and the sections below show by how much. For what the index holds and how it is weighted, see what the S&P 500 is; the stocks hub covers the wider market.
| Measure | Value |
|---|---|
| Easing cycles identified | 9 (1984, 1987, 1989, 1995, 1998, 2001, 2007, 2019, 2024) |
| Average price return 12 months after the first cut | 8.7% |
| Cycles with the S&P 500 higher 12 months later | 7 of 9 |
| Latest month of index data | Oct 2026 |
Key findings
- Usually higher a year later. The index's price rose over the 12 months after the first cut in 7 of 9 cycles. The median gain was 11.3% and the average 8.7%.
- A sharp split on recessions. In the 5 cycles with no recession within 2 years, every 12-month return was positive, and the averages were 6.5% after 3 months, 9.9% after 6 and 16.7% after 12. In the 4 cycles a recession followed, the same averages were 0.6%, −0.3% and −1.2%.
- Losses only when the recession came fast. The two declines, 14.6% after the January 3, 2001 cut and 18.7% after the September 18, 2007 cut, came in cycles where the recession began 2 and 3 months after the cut.
- Not every recession meant a loss. The index was up 11.3% a year after the June 1989 cut, with the recession still a month away, and up 17.1% a year after the August 2019 cut, by which time the two-month 2020 recession had ended.
- Range. The strongest year followed the September 29, 1998 cut: 29.2% in price and 30.8% with dividends. The weakest followed September 18, 2007: −18.7% and −17.0%.
- Dividends. Reinvesting them added 1.1 to 4.9 points to the 12-month return, most in the cycles of the 1980s.
How was the study built?
It finds the first cut of each easing cycle since 1982 and tracks the S&P 500 from that month's average to the averages 3, 6 and 12 months later.
An easing cycle, for this study, opens with a cut of at least a quarter point in the federal funds target that comes after a hike of at least a quarter point. It counts only if the Fed then cuts at least three quarters of a point in total before its next hike. Smaller moves are ignored when deciding which way policy was heading, because in the 1980s the target often moved by an eighth or a sixteenth of a point. The series starts on September 27, 1982, and from December 16, 2008, when the Fed began setting a range rather than a single rate, the study follows the range's upper end.
| Measure | How it is calculated |
|---|---|
| First cut | A cut of at least 0.25 point that follows a hike of at least 0.25 point and begins at least 0.75 point of cuts before the next hike |
| Target before and after | The federal funds target (DFEDTAR) through December 15, 2008, then the upper end of the target range (DFEDTARU) |
| Date of the cut | The first day the new target appears in FRED's daily series |
| Price return after 3, 6 and 12 months | Shiller's S&P 500 average for the month 3, 6 or 12 months after the cut, divided by the average for the month of the cut, minus 1 |
| Total return after 12 months | The same window, with one-twelfth of Shiller's annual dividend rate added and reinvested each month |
| Recession | "Began" when an NBER business-cycle peak month falls within 24 months after the month of the cut, otherwise "None within 2 years" |
| Averages, medians and share higher | Simple averages and medians of the cycle results in each group, and the share of cycles with a positive 12-month price return |
Shiller's file gives one price per month, which it describes as the monthly average of daily closing prices, and dividends computed from S&P's four-quarter totals with linear interpolation to monthly figures. Recession starts come from FRED's NBER-based indicator, USREC, which runs through September 2026 in this snapshot. The study dates a recession from NBER's peak month, the last month of the expansion; NBER itself counts the contraction from the month after. All returns are in nominal US dollars, before costs, taxes and inflation.
Which cycles the rule picked out
Run on FRED's target series as published on November 2, 2026, the rule finds nine cycles. This table comes from the same series but is not part of the downloadable dataset; it shows how far each cycle cut before the next hike of a quarter point or more.
| First cut in FRED's series | Target before and after | Lowest target before the next hike | Total cut | Next hike |
|---|---|---|---|---|
| Sept 20, 1984 | 11.50% to 11.25% | 5.875% | 5.625 points | Apr 30, 1987 |
| Nov 4, 1987 | 7.3125% to 6.8125% | 6.50% | 0.8125 point | Mar 30, 1988 |
| June 6, 1989 | 9.8125% to 9.5625% | 3.00% | 6.8125 points | Feb 4, 1994 |
| July 6, 1995 | 6.00% to 5.75% | 5.25% | 0.75 point | Mar 25, 1997 |
| Sept 29, 1998 | 5.50% to 5.25% | 4.75% | 0.75 point | June 30, 1999 |
| Jan 3, 2001 | 6.50% to 6.00% | 1.00% | 5.50 points | June 30, 2004 |
| Sept 18, 2007 | 5.25% to 4.75% | 0.25% | 5.00 points | Dec 16, 2015 |
| Aug 1, 2019 | 2.50% to 2.25% | 0.25% | 2.25 points | Mar 17, 2022 |
| Sept 19, 2024 | 5.50% to 5.00% | 3.75% | 1.75 points | Sept 17, 2026 |
The 0.25% lows of the 2007 and 2019 cycles are the top of a 0 to 0.25% range. Two judgment calls show up in the table. The target rose by an eighth of a point three times between January and March 1985, too little to count as a hike under the rule, so the cuts of 1985 and 1986 belong to the cycle that began in September 1984 instead of starting a new one. And later cuts never open a cycle of their own: the emergency cuts of March 2020 sit inside the cycle that began in August 2019, and the three cuts of late 2025 inside the one that began in September 2024. The quarter-point hike decided on September 16, 2026 closed that last cycle, and the target range has been 3.75% to 4.00% since.
Decision dates, effective dates and the pre-1994 problem
Each cut is dated by the first day the new target appears in FRED's daily series, which is not always the day the Federal Open Market Committee (FOMC) decided.
| Cycle | Date used | How the decision reached the public |
|---|---|---|
| 1984 | Sept 20, 1984 | Not announced; the date comes from a 2005 reconstruction |
| 1987 | Nov 4, 1987 | Not announced; reconstructed. The cut came 16 days after the October 19, 1987 crash |
| 1989 | June 6, 1989 | Not announced; reconstructed |
| 1995 | July 6, 1995 | Announced that day by Chairman Alan Greenspan, the first announcement to give the funds rate level |
| 1998 | Sept 29, 1998 | FOMC statement the same day |
| 2001 | Jan 3, 2001 | Decided on a conference call between scheduled meetings and announced the same day |
| 2007 | Sept 18, 2007 | FOMC statement the same day |
| 2019 | Aug 1, 2019 | Decided and announced on July 31, 2019 |
| 2024 | Sept 19, 2024 | Decided September 18, 2024; the implementation note made the new range effective September 19 |
Since 2008 the new range has usually appeared in FRED the day after the decision. Because returns are measured by calendar month, that matters only when the extra day crosses a month end, and in 2019 it did: the July 31 decision counts as an August cut, so August's average is the starting point.
The three earliest dates have a different weakness. Before February 4, 1994, the FOMC did not announce its decisions, and its minutes for that meeting describe the prompt announcement as "a departure from past Committee practice." FRED's notes say its values for 1982 to 1993 come from a 2005 St. Louis Fed working paper by Daniel Thornton that rebuilt the target series from FOMC records. A 2003 Fed staff history of FOMC communication adds that for much of the 1980s the Committee steered borrowed reserves rather than the funds rate, and that its operating targets were "not so apparent to outside observers" in those years. The 1984, 1987 and 1989 dates therefore mark when the Fed's internal target changed, as reconstructed later, not when investors learned of it.
How did the S&P 500 do after each first cut?
The index rose over the following year in seven cycles and fell in two, from 29.2% after the 1998 cut to −18.7% after the 2007 cut.
| First cut | Target before | Target after | Price return after 3 months | After 6 months | After 12 months | Total return after 12 months | Recession |
|---|---|---|---|---|---|---|---|
| Sep 20, 1984 | 11.5% | 11.3% | -1.0% | 8.0% | 10.8% | 15.7% | None within 2 years |
| Nov 4, 1987 | 7.3% | 6.8% | 5.3% | 4.5% | 10.6% | 14.5% | None within 2 years |
| Jun 6, 1989 | 9.8% | 9.6% | 7.3% | 7.7% | 11.3% | 15.0% | Began Jul 1990, 13 months later |
| Jul 6, 1995 | 6.0% | 5.8% | 4.6% | 10.2% | 15.6% | 18.2% | None within 2 years |
| Sep 29, 1998 | 5.5% | 5.3% | 16.6% | 25.6% | 29.2% | 30.8% | None within 2 years |
| Jan 3, 2001 | 6.5% | 6.0% | -10.9% | -9.8% | -14.6% | -13.5% | Began Mar 2001, 2 months later |
| Sep 18, 2007 | 5.3% | 4.8% | -1.2% | -12.0% | -18.7% | -17.0% | Began Dec 2007, 3 months later |
| Aug 1, 2019 | 2.5% | 2.3% | 7.2% | 13.1% | 17.1% | 19.3% | Began Feb 2020, 6 months later |
| Sep 19, 2024 | 5.5% | 5.0% | 6.9% | 1.1% | 17.1% | 18.6% | None within 2 years |
Reading along a row shows how each year unfolded. After the September 2007 cut the index was only 1.2% lower after 3 months, but 12.0% lower after 6 and 18.7% lower after 12, as the recession that began in December 2007 deepened. After the January 2001 cut the fall came sooner, 10.9% within 3 months. The 1984 row went the other way: a 1.0% dip after 3 months, then gains of 8.0% and 10.8%. In 1987 and 2024 the index gave back part of an early gain by the 6-month mark and still ended the year higher: 5.3%, 4.5% and 10.6% in the first case, and 6.9%, 1.1% and 17.1% in the second.
The strongest cycle, 1998, led at every horizon, with 16.6% after 3 months, 25.6% after 6 and 29.2% after 12. The Fed said that day it was easing "to cushion the effects on prospective economic growth in the United States of increasing weakness in foreign economies and of less accommodative financial conditions domestically." The statements that opened the two losing cycles described strains in the US economy itself: the January 2001 cut cited "further weakening of sales and production," and the September 2007 statement warned that tighter credit "has the potential to intensify the housing correction."
Starting conditions varied widely. The target stood at 11.50% before the 1984 cut and at 2.50% before the 2019 cut. The first move was a quarter point in five cycles and half a point in four. Every quarter-point opener was followed by a higher index a year later; the half-point openers include both losses, after 2001 and 2007, and two gains, 10.6% after 1987 and 17.1% after 2024. Nine cycles are too few to tell whether the size of the first cut carried information of its own or mostly reflected how worried the Fed already was.
Did it matter whether a recession followed?
Yes: the 12-month average was 16.7% in cycles with no recession within 2 years and −1.2% in cycles where one began.
| Cycles | Number | Average after 3 months | Average after 6 months | Average after 12 months | Median after 12 months | Share higher after 12 months |
|---|---|---|---|---|---|---|
| All cycles | 9.00 | 3.9% | 5.4% | 8.7% | 11.3% | 78.0% |
| Recession began within 2 years | 4.00 | 0.6% | -0.3% | -1.2% | -1.6% | 50.0% |
| No recession within 2 years | 5.00 | 6.5% | 9.9% | 16.7% | 15.6% | 100.0% |
The gap is 17.9 percentage points at 12 months, 16.7% against −1.2%, and it shows at the shorter horizons too: 6.5% against 0.6% after 3 months and 9.9% against −0.3% after 6. The medians point the same way, 15.6% against −1.6%, and so does the share of cycles with a higher index a year on, 100% against 50%.
Four cases make a fragile average. The recession group holds gains of 17.1% after 2019 and 11.3% after 1989 alongside the two weakest results in the table, −14.6% after 2001 and −18.7% after 2007. What separated them in this sample was timing. In 2001 and 2007 the recession began 2 and 3 months after the first cut, so most of the 12-month window fell inside it. In 1989 the window closed in June 1990, a month before the July 1990 peak. In 2019 it ran to August 2020, past the end of the two-month recession that NBER dates from a February 2020 peak to an April 2020 trough.
The all-cycle average describes neither group well. At 8.7%, it sits 8.0 points below the no-recession average and 9.9 points above the recession average, which is why this study leads with the split rather than a single headline number.
The split was only visible later
The grouping uses hindsight. NBER's dating committee names a peak only once it is confident a recession has occurred, and its announcements for the four peaks in this study came months afterward: the July 1990 peak on April 25, 1991, the March 2001 peak on November 26, 2001, the December 2007 peak on December 1, 2008 and the February 2020 peak on June 8, 2020. On the day of each first cut, the row a cycle would end up in was not yet knowable.
How much did dividends add?
Between 1.1 and 4.9 percentage points over 12 months, with the largest additions in the 1980s cycles and the smallest after the 2001 cut.
| First cut | Price return, 12 months | Total return, 12 months | Added by dividends |
|---|---|---|---|
| Sept 1984 | 10.8% | 15.7% | 4.9 points |
| Nov 1987 | 10.6% | 14.5% | 3.9 points |
| June 1989 | 11.3% | 15.0% | 3.7 points |
| July 1995 | 15.6% | 18.2% | 2.6 points |
| Sept 1998 | 29.2% | 30.8% | 1.6 points |
| Jan 2001 | −14.6% | −13.5% | 1.1 points |
| Sept 2007 | −18.7% | −17.0% | 1.7 points |
| Aug 2019 | 17.1% | 19.3% | 2.2 points |
| Sept 2024 | 17.1% | 18.6% | 1.5 points |
The difference is roughly a year of dividends as a share of the starting price, plus a little from reinvesting them, so it was largest when the index's dividend yield was high. It ran from 3.7 to 4.9 points in the three cycles of the 1980s and from 1.1 to 2.6 points from 1995 on. Dividends softened the losing years only slightly: after the 2007 cut, they turned an 18.7% price fall into a 17.0% total loss. How reinvestment works in an account is covered in dividend reinvestment.
Worked example: $10,000 through the best and worst years in the table
Multiply the starting amount by one plus each return.
- After the September 1998 cut, price only: $10,000 × 1.292 = $12,920. With dividends reinvested: $10,000 × 1.308 = $13,080.
- After the September 2007 cut, price only: $10,000 × 0.813 = $8,130. With dividends reinvested: $10,000 × 0.830 = $8,300.
Both pairs use monthly averages, so they describe someone who bought and sold at roughly the average level of each month, before fund costs and taxes.
How much does the monthly-average method change the numbers?
Enough to move a 12-month result by several points: for the 2019 cycle, daily closes give 9.8% to 10.8% against the study's 17.1%.
Shiller's file has one price per month, so each return compares the average level of one month with the average level of another. That smooths out single-day jumps, but it also blurs exactly when the cut happened: a cut on the first trading day of a month and one on the last share the same starting value. FRED's daily S&P 500 series, published by S&P Dow Jones Indices and available on FRED from October 31, 2016, makes it possible to check the two most recent cycles.
Worked example: Rebuilding the 2024 row from daily closes
Average the daily closes in each month from FRED's SP500 series, then divide.
- September 2024, the month of the cut: an average of 5,621.26 over 20 trading days
- 3 months later, December 2024: 6,010.91 ÷ 5,621.26 − 1 = 6.9%
- 6 months later, March 2025: 5,683.98 ÷ 5,621.26 − 1 = 1.1%
- 12 months later, September 2025: 6,584.02 ÷ 5,621.26 − 1 = 17.1%
All three match the study's row, which is built from Shiller's monthly averages.
Measuring from single closes gives different answers, and the choice of starting day matters as much as the method.
| Starting point | 2019 cycle: 3, 6 and 12 months | 2024 cycle: 3, 6 and 12 months |
|---|---|---|
| Study: average of the month of the cut | 7.2%, 13.1%, 17.1% | 6.9%, 1.1%, 17.1% |
| Close on the decision day (July 31, 2019; Sept 18, 2024) | 1.9%, 8.2%, 9.8% | 4.5%, −0.1%, 18.0% |
| Close on the first day of the new target (Aug 1, 2019; Sept 19, 2024) | 3.8%, 9.2%, 10.8% | 2.7%, −0.7%, 16.6% |
| Average of the decision month (July 2019) | −0.6%, 9.4%, 7.1% | Same as the study |
Close-to-close returns run to the last close on or before the same date 3, 6 and 12 months later, from FRED's SP500 series as retrieved on November 2, 2026. They check the method and are not part of the downloadable data.
In 2019 both ends of the window pull the monthly figure up. The August 2019 average, 2,897.50, sat below the July 31 close of 2,980.38, and the August 2020 average, 3,391.71, sat above the July 31, 2020 close of 3,271.12. Had the cut been dated by its July 31 decision, July would have been the starting month and the 12-month figure about 7.1% instead of 17.1%. In 2024 the methods agree within a point at 12 months, but the starting day alone moves the 3-month figure from 4.5% to 2.7%, because the index rose 1.7% on September 19, 2024, the day after the decision.
What are the limits of this study?
The main limits are the small sample, recession labels drawn from hindsight, monthly averaging, reconstructed early dates and price-only headline returns.
- Nine cycles, two small groups. The recession split compares 4 cycles with 5, so a single cycle carries a quarter or a fifth of its group's average; the 1998 cycle's 29.2% is 12.5 points above its group's 16.7%. Slicing nine cases by group and by three horizons also makes it easy to find a pattern that would not repeat.
- Hindsight. The recession column uses peaks that NBER named months later, so it describes what happened rather than what was known at the first cut.
- A provisional last row. The 2024 cycle's 2-year window ended in September 2026, the last month of USREC in this snapshot. NBER's most recent announcement, dated July 19, 2021, set the April 2020 trough, and it has named no peak since. A peak dated later inside that window would move the 2024 cycle into the recession group.
- Monthly averages. Every return is approximate and depends on where in the month the cut fell, as the 2019 check shows.
- Reconstructed early dates. The 1984, 1987 and 1989 dates come from a 2005 rebuild of the target series, covering years when the Fed did not announce its decisions.
- Rule choices. A different cycle rule would change the rows. Ignoring moves under a quarter point folds the 1985 and 1986 cuts into the 1984 cycle, and measuring only from the first cut says nothing about returns after the later cuts in long cycles such as 2001 and 2007.
- Price, nominal and before costs. The main columns leave out dividends, and no figure is adjusted for inflation, fund fees or taxes; real returns and inflation shows how to convert a nominal return.
- One market. The study covers the large US companies in the S&P 500 only; bonds, smaller companies and markets outside the US are beyond its scope.
What does this history say about rate cuts and stocks?
In these nine cycles, the year after a first cut went well or badly in line with whether a recession arrived soon afterward.
Past performance never predicts future results, and nine episodes cannot say which kind of cycle comes next. What the table can do is put a common claim in context. "Stocks rise after the Fed starts cutting" held in 7 of the 9 cycles here, but it leaves out the two cycles that lost 14.6% and 18.7%, and the fact that what set those years apart became clear mainly in hindsight.
Questions to ask of any 'stocks after rate cuts' statistic
- Which cycles are counted, and under what rule?
- Is the return measured from the first cut only, or from every cut?
- Does it start at the decision, the effective date or a monthly average?
- Is it a price return or a total return with dividends?
- Are recessions labelled with what was known at the time, or with NBER dates set later?
- How many cycles sit behind each average?
A rate cut also changes the other side of a portfolio, because yields on cash and Treasury bills follow the Fed down; how Fed rate decisions affect T-bill rates shows how quickly. For what deep declines have looked like and how long recoveries took, see bear markets and corrections. Investing a sum all at once or in stages is weighed in lump sum vs dollar-cost averaging, and the DCA calculator backtests regular purchases on recent prices. The SPY page shows the latest price of a large S&P 500 fund, and bitcoin vs the S&P 500 compares the index's last decade with a far more volatile asset. A separate study of yield curve inversions covers a signal often watched for recession risk.
How to cite and reuse the rate-cut data
The rows are free to download and reuse under the CC BY 4.0 license, as long as the source is credited with a link to this page.
- The CSV file holds the main table, one row per cycle. The summary and key numbers can be quoted with the same credit.
- Give each figure its cycle and horizon, for example: the S&P 500 rose 17.1% in price in the 12 months after the September 19, 2024 cut, measured between monthly averages.
- Say that the headline returns are price-only and based on monthly averages, and that the 2024 cycle's recession label could change if NBER dates a new peak.
- Suggested credit: Source: Sonic Wallet analysis of Robert Shiller's U.S. stock market data and FRED series DFEDTAR, DFEDTARU and USREC, index data through October 2026.
The editorial policy explains how studies are checked, dated and corrected.